The log power meter can provide measurements from less than 70 dBm (0.1 nanowatt) to over 13 bBm (20 mW). Please refer to the QST article for further details on its use and capabilities.

I deviated a bit from the standard construction of the power meter as described in the article, taking to heart Wes's emphasis on shielding on his web page. I built a separate shielded enclosure for the front end circuitry, including the AD8307 log amp IC, and I fed power into this shielded enclosure, "the log-amp cabin" via a 1,000pF feedthrough cap, and fed the signal out via another of the same. The only other penetration to the shielded environment is the BNC connector. I'll describe how I built the power meter in the following paragraphs, with photos of each step.

Click on the photos for a higher resolution image.

I used 1/32" double-sided pc board stock for the log amp enclosure. The basic pieces are the front and rear panels, the floor and one side. Later, I'll show how I used tin for the other parts of the enclosure.

The BNC connector is used to mount the front of the enclosure to the front panel. As you can see, I've applied tin strips around the front. To make sure that the BNC connector contacts the front panel on both sides, I used an extra-long BNC bulkhead connector, and I put an extra nut between the front panel and the front tinned copper section of the log amp enclosure.

I designed a printed circuit board for the components not on the input/log amp board. This includes the LM358 op amp and the 5-Volt regulator circuits. I made the right-angle mounting brackets from some thin steel that I cut with my tin snips.

I used a stepped drill bit to drill the holes, and enlarged the hole for the meter with a rotary rasp mounted on the drill. This photo show the DVM atttachment point, which happens to be another type of 1,000 pF feedthrough capacitor. I think I'll be able to construct a DVM probe plug that will mate coaxially with the cap.

The view of the cabinet from the rear.

I didn't want to solder to the AD8307 loag amp chip dead-bug style, so I made a Manhattan-ish pad to use with a DIP socket. The pad is to be soldered to the board, instead of being glued, so I provided extra solder points at both ends of the board. My concern was that glue might not adhere well to the tinned board surface, and that the heating involved with putting on the tin roof and wall might pop it loose. That happened to a low pass filter I built using the pc board and tin method.

Photo of the completed log amp IC pad, tinned with solder, andchecked for shorts. The dark color is an artifact due to the camera flash.

The IC socket soldered to the pad.

I drew up a rough to-scale layout plan for the board to ensure that I would have room to fit the input components between the BNC connector and the IC pad. You can see in the photo the grounding/pad position-locking component leads that I attached to the IC pad. The feed-through caps are soldered into their holes in the side wall.

I checked the board for fit into the front log amp housing. I also had installed the capacitors that go from the IC socket pins to ground. Staged for installation are the front end resistors and the capacitor-inductor combo, C1 and L1.

I tacked the log amp board in place and installed the front end components. I used a 1% 51.1 ohm resistor for R1, soldered from the BNC center post to ground. The RF input series path includes the parallel combination of R2 (470 ohms) and C2 (15 pF) and L1/C1, connected to pin 8 of the log amp. By the way, prior to finally boxing the board up with its tin sheathing, I rechecked L1 with a 3/16" drill bit for size.

Before I went much further, I figured I should label the feed through caps on the outside of the wall. Otherwise, based on experience, after soldering everything shut, I'd surely have some anxiety over whether the power and DC signal current output feedthroughs were the right ones!

The side panel soldered in place to the front box.

This was the final good opportunity to insert the AD8307 log amplifier IC in its socket. Note that I used a wrist grounding strap while doing this.

I used a paper template to determine the size and bend points of the tin sheeting to be used for the top and right side of the "log amp cabin."

Side and roof panels soldered in place.

Since I used double-sided pc board, I could not effectively solder the outside of the joint and get RF-tightness; therefore, I made a trim piece out of tin, and soldered it over the remaining edges of the pc board. The "hobby tin" sheets, an Ace Hardware stock item, are soft and easy to bend, and they accept solder readily.

The "cabin" completely installed. The next steps were to connect the wiring for the power from the 9 Volt battery snaps to the switch and power LED and the main circuit board. Then, the circuit board connections were made with 5 Volt power to the log amp, the log amp signal, the auxiliary DVM connection and the panel meter. I used a 2.2k dropping resistor for the orange LED to make it bright enough to be visibly on, but to save on the current draw.

The insides of the completed power meter. Pretty simple, actually, considering what this device is capable of.

A shot of the bottom foil of the main PC board. Note the credit to the original circuit designers, W7ZOI and W7PUA. I put this information on all of my PCBs.

Wiring detail showing component side of the PC board. Probably overkill for what is a DC signal at this point, but I used RG-174 to route the signal from the log amp circuit's feed through cap to the main board.

The completed log power meter. Due to my choice of the location of the 9 Volt battery holder, I guess I'll have to put the calibration card in a frame on top of the meter.

The meter, powered up. It shows a background reading of "1.5 Volts"...actually, since this meter is hooked up as milliammeter with a full scale reading of 1 mA, the reading is .1 mA. At this point I haven't checked the calibration of the meter, but looking at the calibration chart in the QST article, mine is consistent with that one, the reading being less than -70 dBm at this point.